In [1]:
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
%matplotlib inline
In [2]:
df1 = pd.read_csv('mtcars.csv')
print(df1.head())
print(df1.shape)
Line Plot¶
In [3]:
#plt.style.use('fivethirtyeight')
plt.style.use('classic')
x1 = np.arange(0,25)
#print(x1)
y1 = x1 * 2
#print(y1)
y2 = x1 ** 2
plt.plot(x1,y1, color='yellow', linestyle='dashed', marker='o', label='linear plot')
plt.plot(x1,y2, color='blue', linestyle='solid', marker='>', label='expo plot')
plt.plot([1]*600,list(range(600)), color='#ff00ff', linestyle='dotted', marker='>', label='expo plot')
plt.xlabel('x_label')
plt.ylabel('y_label')
plt.title("first plot")
plt.legend(loc=0)
plt.savefig("myfirstplot.png")
plt.show()
In [4]:
plt.figure(figsize=(10,8),dpi=50)
plt.plot(x1,y1, color='yellow', linestyle='dashed', marker='o', label='linear plot')
plt.savefig("myfirstplot2.png",dpi=200)
In [5]:
df1.sort_values(by='mpg').plot(x='wt',y='mpg', kind='line')
Out[5]:
In [6]:
df2 = df1.sort_values(by='mpg')
plt.plot(df2.wt, df2.mpg)
#plt.plot('wt','mpg', data=df2)
Out[6]:
Scatter Plot¶
In [7]:
plt.scatter(x1,y2)
Out[7]:
In [8]:
plt.scatter(df2.wt,df2.mpg)
Out[8]:
In [9]:
plt.scatter(df1.gear, df1.mpg)
Out[9]:
In [10]:
df1.gear.value_counts()
Out[10]:
In [11]:
df1.plot(kind='scatter', x='gear',y='hp')
Out[11]:
Bar Plot¶
In [12]:
stockname = ['goog', 'ibm', 'tcs','hll']
stockvalues = [1000,2000,3000, 4000]
plt.bar(stockname,stockvalues, color='rg', width=0.75,align='center')
Out[12]:
In [13]:
plt.barh(stockname,stockvalues, color='rg')
Out[13]:
In [14]:
vc = df1.gear.value_counts()
print(vc)
vc.plot(kind='bar')
for index,values in enumerate(vc.values):
print(index,values)
plt.text(index,values,str(values))
Histogram¶
In [15]:
agevalues = np.random.randint(0,101,size=100)
print(agevalues)
plt.hist([agevalues,agevalues*2], bins=10, rwidth=0.9,stacked=False,color=['r','g'])
Out[15]:
In [16]:
plt.hist([agevalues,agevalues*2], bins=10, rwidth=0.9,stacked=True)
Out[16]:
In [17]:
df1.plot(kind='hist', y='mpg')
Out[17]:
In [18]:
df1.plot(kind='kde', y='mpg')
Out[18]:
Box Plot¶
In [19]:
plt.style.use('ggplot')
v1 = list(range(100,200))
v1.append(10)
v1.append(300)
v1.append(400)
plt.boxplot(v1)
plt.show()
In [20]:
df1.hp.plot(kind='box')
Out[20]:
In [21]:
df1[df1.hp > 300]
Out[21]:
Pie Chart¶
In [22]:
vc = df1.gear.value_counts()
print(vc)
In [23]:
plt.pie(x=vc.values, labels=vc.index, autopct='%.1f')
print()
In [24]:
plt.pie(x=vc.values, labels=vc.index, autopct='%.1f', explode=[0,0,0.25])
plt.legend(['abc','def','ghi'])
print()
sns Bar Plot¶
In [25]:
sns.barplot([100,200,300],ci=95, estimator=np.mean)
Out[25]:
In [26]:
from seaborn.utils import ci
sns.barplot(x='mpg', data=df1)
x = ci(df1.mpg)
print(x)
x = np.std(df1.mpg)
print(x)
x = np.mean(df1.mpg)
print(x)
In [27]:
sns.barplot([100,200,300,400,500])
Out[27]:
In [28]:
sns.barplot(x='gear',y='mpg',data=df1)
Out[28]:
In [29]:
sns.barplot(x='gear',y='mpg',data=df1, hue='cyl')
Out[29]:
In [30]:
dp1 = df1.pivot_table(index='gear', columns='cyl', values='mpg',aggfunc='count')
print(dp1)
sns.scatter¶
In [31]:
sns.scatterplot(x='gear',y='mpg',hue='cyl',data=df1)
Out[31]:
In [32]:
sns.scatterplot(x='gear',y='cyl',data=df1)
Out[32]:
In [33]:
sns.stripplot(x='gear',y='cyl',data=df1)
Out[33]:
In [34]:
sns.swarmplot(x='gear',y='cyl',data=df1)
Out[34]:
Regression Plots¶
In [35]:
sns.regplot(data=df1, x='wt', y='mpg')
plt.title('regplot')
Out[35]:
In [36]:
sns.lmplot(data=df1, x='wt', y='mpg')
Out[36]:
sns Distribution¶
In [37]:
sns.distplot(df1.mpg, bins=5, hist=True, kde=True, rug=True, hist_kws={'rwidth':0.5, 'color':'y'})
Out[37]:
sns IQR plot¶
In [38]:
sns.boxplot(df1.mpg)
Out[38]:
In [39]:
sns.boxplot(df1.mpg, orient='v')
Out[39]:
In [40]:
sns.boxplot(data=df1,x='gear', y='mpg')
Out[40]:
In [41]:
sns.boxplot(data=df1,x='gear', y='mpg', hue='am')
Out[41]:
In [42]:
sns.boxenplot(data=df1,x='gear', y='mpg', hue='am')
Out[42]:
In [43]:
sns.violinplot(data=df1,x='gear', y='mpg', hue='am')
Out[43]:
Multi plots¶
In [44]:
sns.pairplot(df1[['mpg', 'gear','cyl']])
Out[44]:
In [45]:
sns.jointplot(data=df1, x='gear', y='mpg',marginal_kws={'bins':10, 'kde':True,'color':'y'}, joint_kws={'color':'b'})
Out[45]:
In [46]:
pd.set_option('display.width',1000)
#heat map
cmatrix = df1.corr()
cmatrix
Out[46]:
In [47]:
plt.figure(figsize=(16,12))
sns.heatmap(cmatrix, annot=True)
Out[47]:
Subplotting¶
In [48]:
fig, ax = plt.subplots(3,3, figsize=(10,12),dpi=100)
ax[0][0].set_title('00')
ax[0][1].set_title('01')
ax[1][0].set_title('10')
ax[1][1].set_title('11')
ax[0][0].set_xlabel('x00')
ax[0][1].set_xlabel('x01')
ax[1][0].set_xlabel('x10')
ax[1][1].set_xlabel('x11')
#print(type(ax[0][0]))
ax[0][0].scatter(df1.gear, df1.mpg)
ax[0][1].hist(df1.mpg)
ax[1][0].bar(df1.cyl, df1.mpg)
ax[1][1].scatter(df1.carb, df1.mpg)
plt.show()
In [49]:
print(ax)
In [50]:
print(fig)
In [51]:
ax[0][0].set_title('hello')
Out[51]:
Plotly¶
In [52]:
import plotly
plotly.offline.init_notebook_mode()
import plotly.express as px
import plotly.graph_objects as go
In [53]:
fig = px.box(data_frame=df2, y='mpg')
fig.show()
In [54]:
fig = px.scatter(data_frame=df1, x='cyl',y='mpg')
fig.show()
In [55]:
v1 = list(range(100,200))
v1.append(10)
v1.append(300)
v1.append(400)
fig = px.box(y=v1)
fig.show()
In [56]:
fig = px.histogram(data_frame=df1, x='mpg', nbins=10, title='histplot')
fig.show()
In [57]:
px.histogram(data_frame=df1, x='hp', nbins=10, title='histplot')
In [ ]: